|
||||||||
|
CHRONIC DISEASES Sickle Cell Disease Cured in Mouse ModelReaching a milestone in the advance against genetic disorders, a consortium led by Philippe Leboulch, HMS assistant professor of medicine at Brigham and Women's Hospital, has used gene therapy to cure a model of sickle cell disease in mice. The achievement, announced in the Dec. 14 Science, marks a breakthrough in what initially seemed like a simple task: to cure a rather straightforward genetic defect by introducing a new gene into the stem cells of bone marrow. The effort proved more arduous than expected, forcing scientists to confront the complexities of stem cell biology, virology, gene expression, and protein biochemistry.
After many refinements, Philippe Leboulch, left, and colleagues Robert Pawliuk and Karen Westerman have succeeded in creating a gene therapy that cures all symptoms of sickle cell disease in mouse models. Photo by Graham Ramsay Sickle cell disease has a history of firsts in medical science. In 1957, it became the first genetic disease characterized at the molecular level. In 1976, it also became the first disease, with the related disorder beta-thalassemia, to be diagnosed prenatally. With its position at the forefront of genetic understanding, scientists naturally hoped that it also would be the first disease to be cured by a genetic therapy. In 1979, the first gene therapy experiment for sickle cell disease was attempted, but more than 20 years and many stumbles and false starts later, a genetic therapy for use in humans still does not exist. Error AmplificationSickle cell disease is caused by a single point mutation in the beta-globin gene that results in the substitution of one amino acid. This small error is enough to change the properties of the protein: when sickled, hemoglobin dumps its oxygen in tissues, and it tends to stick to itself, polymerizing into chains that render red blood cells abnormally rigid, adhesive, and distorted into the hallmark sickle shapes that characterize the disease. The simplicity of the sickle cell mutation made it seem a perfect candidate for gene therapy. The cell population was also accessible, since bone marrow could be harvested and a potential antisickling gene could be introduced ex vivo. But the construct that Leboulch and his colleagues have created is not the conceptually straightforward process initially envisioned but a piecemeal affair, representing years of discovery and subtle manipulations.
The red blood cells of untreated sickle cell mice (top) have a characteristic crescent shape at lower oxygen tension, which stimulates red blood cell sickling in vitro. In contrast, treated sickle cell mice (bottom) have normal red blood cells. Courtesy of Philippe Leboulch. The first problems encountered were how to package and express the gene of choice. Of the panoply of available vectors, only retroviral carriers seemed capable of allowing permanent gene transfer to bone marrow stem cells and their progeny. Initial attempts using beta-globin cDNA achieved very little expression, so it was thought that the entire gene would be more effective. However, because retroviral vectors are RNA viruses, it was likely that the introns of the beta-globin gene would be excised before it could be packaged into virions. A solution to the problem, achieved by Richard Mulligan, now the Mallinckrodt professor of genetics at Children's Hospital, was to place the gene backwards, rendering the splice sites gibberish to the splicing machinery and fooling it into copying the entire gene, introns and all. By attaching the beta-globin promoter, expression could be tissue specific. Further progress came when a group at MIT, led by Irving London, the Grover M. Hermann professor emeritus of health sciences and technology in the Harvard-MIT Division of Health Sciences and Technology, discovered a region of DNA upstream of the gene that contains strong globin-enhancer elements. It was hoped that better expression could be achieved by including portions of this sequence, dubbed the locus control region, into the vector. But the resulting constructs were extremely unstable with multiple rearrangements. In the mid-'90s, Leboulch discovered that just as people sometimes mistakenly hear messages in a record played backwards, reversing the gene and adding control region elements not normally transcribed had created cryptic splicing sequences that caused the instability. His team identified and painstakingly mutated these sites to prevent unwanted splicing. With these enhancements, Leboulch said, "We were able to have long-term reconstitution in mice, and we had long-term expression, but the titer was still low." Deciding he needed a way to fit larger portions of the locus control region into the vector to get higher expression levels, Leboulch proposed using an export element, such as the HIV Rev/RRE system, which efficiently exports unspliced, full-length viral RNA to the cytoplasm. Another group was able to build a stable HIV virus containing larger sequences of the locus control region with the help of these elements and to ameliorate beta-thalassemia in mice, but more work was required to affect sickle cell disease. Antisickling EdgeAfter resolving how to package and express the gene, doubt still remained whether the gene of choice was the correct one. The normal beta-globin gene inhibits sickle hemoglobin polymerization only at very high concentrations by a mere dilution effect, whereas other globins, such as fetal gamma-globin, can actually inhibit sickling through specific protein-protein contacts. But attempts to introduce gamma-globin were disappointing. Leboulch's team wondered if they could create a beta-globin variant with all the antisickling properties of gamma-globin. Knowing that there were only minor differences between beta- and gamma-globin, they reviewed the literature to determine which amino acids gave gamma-globin the antisickling edge and altered the beta-globin gene to include one that seemed a likely candidate.With the vector construct optimized through years of refinements, the next question was how to test it. Because sickle cell disease does not exist in mice and the endogenous mouse globin genes inhibit sickling, the team used two mouse models developed by other investigators: one that expresses a variant of human sickle beta-globin that can cause sickling even in the presence of mouse hemoglobin, and another in which all mouse globins are disrupted and only human globin chains, including the sickle globin, are expressed. They transplanted mice with transduced stem cells, and found that up to 99 per cent of red blood cells permanently expressed the antisickling globin at high levels. "We also cured or significantly altered all the symptoms and anomalies we could detect in those mice: irreversible sickle cell counts, anemia, red cell density, spleen enlargement, and a specific urine concentration defect," said Leboulch. Just as the journey to a cure in mice has been tortuous, much work must be done before the therapy can be attempted in humans. The trade-off of using aggressive lentiviral vectors is the risk of contamination with a replication-competent virus. In addition, the mice had to be exposed to a regimen of radiation before bone marrow replacement to eliminate nontransduced stem cells. But although the hopes for gene therapy that scientists once held have been tempered by the sheer biological complexities, the study does support the notion that science can someday yield therapies perfectly constructed to cure disease. --Courtney Humphries |
